Preferred link based delay-constrained least-cost routing in wide area networks

  • Authors:

  • R Sriram;G Manimaran;C.Siva Ram Murthy

  • Affiliations:

  • Department of Computer Science and Engineering, Indian Institute of Technology, Madras 600 036, India;Department of Computer Science and Engineering, Indian Institute of Technology, Madras 600 036, India;Department of Computer Science and Engineering, Indian Institute of Technology, Madras 600 036, India

  • Venue:
  • Computer Communications
  • Year:
  • 1998

Quantified Score

Hi-index 0.25

Visualization

Abstract

Multimedia applications involving digital audio and/or digital video transmissions require strict QoS constraints (end-to-end delay bound, bandwidth availability, packet loss rate, etc.) to be met by the network. To guarantee the real-time delivery of packets satisfying these constraints, a real-time channel (D. Ferrari and D.C. Verma, A scheme for real-time channel establishment in wide-area networks. IEEE JSAC, 8(3), 368-379, 1990) needs to be established before the transmission of packets of a connection can begin. The establishment of such channels requires the development of efficient route selection algorithms that are designed to take into account the QoS constraints. The general problem of determining a least-cost delay-constrained route in a given communication network has been proved to be NP-hard (M.R. Garey and D.S. Johnson, Computers and Intractability: a guide to the theory of NP-completeness, W.H. Freeman, 1979). In this paper, we describe a preferred link approach to distributed delay-constrained least-cost routing in order to establish real-time channels. The approach attempts to combine the benefits of probing and backtracking based algorithms (better adaptiveness and wider search) with the advantages of distance-vector type algorithms (lower setup time). The scheme is flexible in that a variety of heuristics can be employed to order the neighbouring links of any given node. Three heuristics are proposed and their performance is studied through simulation experiments. The simulation results indicate that the proposed heuristics provide better performance than other preferred neighbour methods, in terms of increased call acceptance rate and lower average route cost. The heuristics are also shown to adapt much better to dynamic variations in network and link characteristics.